Red Hat Bugzilla – Bug 64801
changing to console1 from console 7 causes logout or kdm restart
Last modified: 2007-04-18 12:42:32 EDT
From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.0rc2) Gecko/20020510 Description of problem: I use KDE, and kdm, have not tested this with gdm. If I login to KDE and then after a while switch to console 1 with CTRL+ALT+F1, something happens and my original X console (7) is logged out or restarted because I will have a kdm login screen again. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Login to kdm 2. Switch to console 1 3. Wait a while ( over 30 minutes from my testing ) 4. Check back with computer. Actual Results: You will see login screen of kdm. Expected Results: I expect it to be where I left off, logged into kde. Additional info:
Attach your X server log, and config file please.
Created attachment 57064 [details] XFree86 Log
Created attachment 57065 [details] XFree86 config file
This sounds like a VTswitch problem, either with switching out of, or switching into X with Intel i8x0 hardware. If the X server crashes when xdm/kdm/gdm are used, then it will be restarted which is why you are getting a fresh login window. I recommend using runlevel 3 by changing initdefault in /etc/inittab to 3, and starting X by using "startx". Once you switch away from X and observe a crash of the X server, save a copy of the X log file WITHOUT starting up the X server again. Attach that log file to the bug report. Make sure you're attaching the log from the crash, and not a log from the currently running X session.
Does this problem still occur with all of our rpm package updates applied that have been released as erratum to date? In particular our latest kernel RPM?
There is a bug in the i810 driver which the Intel driver author broke the kernel ABI and did not change the version. This causes various weird problems to occur. Our latest kernel contains a hack workaround for this problem that there is no real way of fixing 100%. People whom have properly installed our newest kernel, and have things configured correctly, do not experience these types of problems anymore. I'm assuming you've done so as well due to lack of response to my last two queries, so I'm closing this bug report as fixed. The problem doesn't occur in Red Hat Linux 8.0, so I'm closing with resolution of CURRENTRELEASE.